Rural Resilience Assessments in the Yangtze River Delta Based on the DPSIR Model

Abstract

The Yangtze River Delta (YRD) region, located inside the Yangtze River Basin, functions as a vital ecological and economic area in China, with its natural environment directly impacting human existence. This study seeks to elucidate the spatial and temporal evolution of rural resilience in the Yangtze River Delta region and its underlying mechanisms by establishing a comprehensive assessment framework for rural resilience, thereby offering a scientific foundation and policy guidance for the region’s sustainable development. The research first established the DPSIR (driving force–pressure–state–impact–response) assessment index system. Subsequently, the entropy weighting method and TOPSIS were utilized to assess and rank the rural resilience levels in the Yangtze River Delta region (Shanghai, Jiangsu, Zhejiang, and Anhui) from 2012 to 2022. Ultimately, partial least squares structural equation modeling (PLS-SEM) was employed to examine the intrinsic logical relationships among the five dimensions of the DPSIR framework and to extract conclusions. The study effectively met the goals of SDG 7 (clean water and sanitation), SDG 8 (decent work and economic growth), and SDG 11 (sustainable cities and communities). The research indicated that (1) the resilience level in the Yangtze River Delta region initially declined, then increased, and eventually attained a condition of stabilization. Changes in the “driving force”, influenced by the “response level” and environmental “pressure”, have affected the resilience level of rural areas. There is heterogeneity in the assessment values and ranges of change among provinces, with the “impact” component exhibiting the most substantial evaluation value. The findings yield policy recommendations for the implementation of diverse regional governance, the establishment of connectivity mechanisms, the customization of strategies to address the specific deficiencies of each province, and the systematic enhancement of rural resilience.

1. Introduction

The Yangtze River Basin (YRB), China’s most important ecological and economic zone, is home to the bulk of the country’s population and economic activity, with its ecological and environmental conditions having a direct impact on national ecological security and sustainable development initiatives. Issues in the Yangtze River Delta region have intensified in recent years, and data from the China Environmental Status Bulletin (2023) show that China’s economic growth over recent decades has resulted in significant environmental challenges, particularly water pollution and biodiversity loss [1,2,3,4,5,6]. Water contamination in rural areas has a negative influence on the quality of life for local residents and causes major harm to ecosystems, especially in communities that rely on agriculture and natural resources. Accelerated urbanization has resulted in population expansion and increased traffic, a decline in green areas, an increase in heat sources, and a rise in the occurrence of severe environmental difficulties, such as the urban heat island effect [7,8,9,10,11,12].

Rural resilience research has been a major emphasis in landscape ecology and sustainable development across the world. Recent research themes have focused on climate change and ecological resilience, which study how rural populations adapt to catastrophic weather events (e.g., droughts and floods) and long-term environmental changes (e.g., land degradation) [13,14,15,16]. Adam M. Straub et al. studied the impact of social capital (trust, cooperation, and exclusivity) on the resilience of rural communities in the southern Great Plains of the United States (Cimarron County, Oklahoma; Union County, New Mexico; and Las Animas County, Colorado) to climate change and drought [17]. The ability of rural economic diversification, such as tourism and renewable energy, to withstand market swings is evaluated [18]. Green development research has revealed a trend of multidimensional deepening by constructing an analytical model of “structure-process-function” in agricultural landscapes, thereby providing biodiversity conservation measures (e.g., ecological corridor building) and greening efforts [19,20]. Marie Arndt, a German researcher, meticulously evaluated the impact of agricultural diversification techniques on sustainability and resilience in temperate regions, especially in Europe, using a study of 142 scientific papers [21]. The impact of migration and demographic aging on the social fabric of villages is investigated, as well as the contribution of indigenous knowledge to the development of resilience [22]. Worldwide, Europe and the United States place a stronger focus on systems and innovation [23]. Europe promotes governmental integration and digital technologies [24], whereas North America focuses on community-driven resilience planning [25]. Asia focuses on the balance of growing urbanization. China prioritizes ecological preservation and industrial transformation in its “rural revitalization” agenda [26]. Africa and Latin America focus on resource dependency and conflict resolution [27].

Recent rural resilience research has emphasized interdisciplinary integration, technology-driven techniques, dynamic assessments, and community participation as methodologies. Nabi Moradpour conducted a spatio-temporal assessment of southern Tehran neighborhoods using the fuzzy Delphi approach and K-means clustering [28]. Luc Ampleman argued for the implementation of Grimes’s “Actantial Model” to reproduce resilience narratives of rural communities in order to handle the complexity and transdisciplinary communication issues in sustainability assessments [29]. Wu presented a paradigm for measuring flood resilience by including multi-source evidence, such as the literature, policy documents, and social media data [30]. He created a comprehensive indicator system and performed an empirical analysis using the Yangtze River Economic Belt as a case study to understand regional inequalities and major contributing variables [31]. Wang created a spatio-temporal assessment matrix by combining the 4R model of crisis management with the multidimensionality of urban systems [32].

This study investigates rural resilience development in the Yangtze River Delta region, with the goal of systematically addressing three primary issues by developing a coupled DPSIR-PLS-SEM assessment model to achieve the objectives of SDG 7 (clean water and sanitation), SDG 8 (decent work and economic growth), and SDG 11 (sustainable cities and communities), with an initial focus on reducing the subjectivity inherent in current assessments. Initially, to mitigate the subjectivity of current assessment methodologies, we develop a multi-dimensional quantitative indicator system for DPSIR to objectively evaluate regional rural resilience levels; subsequently, using panel data from 2012 to 2022, we conduct a comprehensive analysis of the spatial and temporal evolution of resilience levels; finally, by pinpointing the resilience deficiencies across various regions, we recommend tailored policy combinations, and further, this study aims to improve the theoretical understanding of the DPSIR subsystem’s nonlinear action mechanism while also providing empirical evidence to help local governments design targeted interventions and facilitate the effective implementation of the Yangtze River Delta’s rural revitalization strategy.

2. Conceptual Framework and Definitions

2.1. DPSIR-PLS-SEM Framework

The driving force–pressure–state–impact–response (DPSIR) model is a systematic analytical framework that defines causal links within the social–ecological system and offers a solid methodological foundation for assessing landscape resilience. The model divides the dynamic changes in the rural system into five coherent components: “driver-pressure-state-impact-response”. The model combines quantitative measurements and qualitative analysis, allowing researchers to identify significant vulnerabilities and design resilience-building methods, making it a crucial tool in rural sustainable development research (Figure 1). Using the DPSIR model to assess the resilience of rural landscapes in the Yangtze River Basin will help define the fundamental elements influencing landscape changes, highlight their interaction processes, and offer a scientific foundation for developing targeted landscape management strategies.

The present use of DPSIR-PLS-SEM in resilience evaluations exemplifies the characteristics of methodological integration. Zhang’s study uses the DPSIR-PLS model to assess the quality of rural habitats in 31 Chinese provinces between 2018 and 2021. It establishes a regional differential defined by “strong in the east and weak in the west” and proposes personalized governance options by studying the causation among socioeconomic factors, ecological constraints, and policy responses [33]. Sun’s study differs considerably in that he constructs a coastal rural ecological resiliency framework using the DPSER (driving, pressure, state, impact, and response) model [34]. Yi’s research uses the DPSIR model and the SDGs to provide a framework for evaluating urban climate resilience, including a unique indicator: the government’s concern for climate policy [35]. Using the DPSIR-PLS-SEM model to assess the resilience of rural landscapes in the Yangtze River Basin can help identify critical elements influencing landscape change, clarify interaction processes, and provide scientific evidence for developing targeted landscape management plans.

2.2. Rural Resilience Assessment

Rural resilience refers to rural regions’ overall ability to absorb disruptions, adapt to changes, restore functioning, and achieve long-term development in the face of external shocks and ongoing pressures, such as environmental, economic, social, and political causes. Holling first defined ‘resilience’ as the ability of ecosystems to survive disruptions while maintaining core functioning [36]. Cumming’s research combines landscape ecology, resilience theory, and social–ecological system research through the concept of spatial resilience, highlighting the impact of spatial variation on a system’s capacity to adapt to change while offering novel theoretical frameworks and practical strategies for sustainable management [37].

The goal of rural resilience evaluations is to statistically assess the vulnerability and adaptability of rural systems in order to assist policy creation and planning. The major goal is to identify the rural system’s weaknesses and strengths, as well as to provide a scientific framework for developing resilience-enhancing activities. The three major assessment models are quantitative, score and questionnaire, and hybrid. Bruneau et al. (2003) proposed a four-dimensional model for measuring community resilience [38]. The four core elements include technological, organizational, social, and economic components, each of which must meet four important criteria: stability, redundancy, intelligence, and speed. In 2012, the Justice Institute of British Columbia in Canada launched the Rural catastrophe Resilience Planning Guide to help rural, remote, and coastal communities assess catastrophe risk and build resilience.

3. Materials and Methods

3.1. Study Area and Data Sources

The Yangtze River Delta (YRD) is one of China’s most economically vibrant and accessible locations, spanning Shanghai, Jiangsu, Zhejiang, and Anhui provinces and covering around 358,000 square kilometers. The population in this area is considerably concentrated, comprising over 235 million or 16.7% of the national population, and has an average density of 660 people per square kilometer, which is 4.5 times more than the national average. Shanghai has the highest population density, with over 4000 people per square kilometer, whereas major cities like Nanjing and Hangzhou retain concentrations of roughly 1500 people per square kilometer [39].

The region’s GDP per capita is CNY 128,000, which is higher than the national average. Residents’ incomes show a gradient distribution, with Shanghai leading the way at CNY 79,600, followed by Zhejiang at CNY 60,300, Jiangsu at CNY 52,500, and Anhui at CNY 37,900, which, while relatively low, is growing at a faster pace. The strong spending power boosts overall retail sales of consumer products to more than CNY 10 trillion or nearly one-fourth of the national total.

The enterprise ecosystem is vast and diverse, with over 15.6 million registered businesses, accounting for 18.3% of the country. The manufacturing industry has a lengthy history, accounting for 42% of the total and include well-known corporations like SAIC, Haier, and Geely. The digital economy is rapidly growing, with its share reaching 23% and internet behemoths like Alibaba and Pinduoduo establishing a footprint. Furthermore, the area is home to nearly 1800 listed corporations, accounting for one-third of the total A-shares, as well as 126,000 foreign-funded enterprises, accounting for 28% of the nation, demonstrating its international nature. The region is home to the headquarters of the world’s top 500 companies, as well as countless small- and medium-sized organizations (SMEs) focusing on specializations and innovations, all of which contribute to a thriving private economy and a complete industrial eco-system. The region exhibits clear industrial synergies, with Shanghai specializing in finance and high-end manufacturing, Jiangsu in advanced manufacturing, Zhejiang in the digital economy, and Anhui in industrial transfer and the development of new industries, thus complementing each other’s strengths and collaboratively establishing a modernized economic system with a global impact [40].

This research examines the Yangtze River Delta region, especially the cities of Anhui, Shanghai, Jiangsu, and Zhejiang. Data from 2012 to 2022 are reviewed, with a focus on the years 2012, 2014, 2016, 2018, 2020, and 2022. Data indicators were obtained from the China Urban Statistical Yearbook, China Rural Statistical Yearbook, China Statistical Yearbook, and Jiangsu Rural Statistical Yearbook; missing variable values were imputed using linear interpolation, and discrepancies between yearbook data were normalized using Z-scores.

3.2. Selection of “Driver-Pressure-State-Impact-Response” Indicators

Table 1. Selection of latent variable factors for the DPSIR model.

Criterion Layer	Factor Layer	Indicator Layer	Observation Data	Unit
Driving Force	Economic Development	Rural Retail Sales	D1	Billion Yuan
	Population Change	Rural Population	D2	10,000 people
		Proportion of Rural Population	D3	%
Pressure	Ecological Damage	Flood Disaster Area	P1	1000 ha
	Agricultural Pollution	Fertilizer Usage	P2	10,000 t
		Pesticide Usage	P3	t
State	Ecological Environment	Ecological Environment	S1	units
		Nature Reserves	S1	Units
	Natural Disasters	Afforestation Area	S2	1000 ha
Impact	Policy and System	Population Benefiting from Improved Water Access	I1	10,000 people
		Village Clinics	I2	units
	Resource Availability	Key Forestry Projects	I3	ha
Response	Technical Response	Biogas Production	R1	10,000 m3
	Ecological Restoration	Soil Erosion Control Area	R2	1000 ha
	Social Response	Natural Disaster Relief	R3	CNY 1000

displays the multiple latent variable components used in this inquiry. The key driving forces are economic development and population change; in rural regions, retail sales serve as a barometer of economic progress, and the vitality of the rural economy has a direct impact on resource utilization and environmental conservation. To analyze population change, indices of the rural population and the ratio of the rural population to the total population are utilized, which illustrate the distribution of urban and rural populations as well as the urbanization process.

Pressure is largely classed as environmental deterioration and agricultural contamination. The flood-affected region shows the direct impact of extreme weather events on rural communities and poses a substantial threat to ecological and agricultural systems. The use of fertilizers and pesticides measures agricultural operations’ environmental effect; excessive fertilizer usage contaminates soil and water, jeopardizing ecosystem integrity and agricultural produce safety.

Status is divided into ecological and natural disasters. Nature reserves are important repositories of biodiversity and ecological benefits. The area allocated for afforestation represents the accomplishment of ecological restoration and forestry progress, and it serves as an important method for improving the ecological environment. The impacted region immediately displays the amount of rural areas affected by natural catastrophes and serves as an important indicator of systemic vulnerability.

Impact is divided into policy systems and resource availability. The policy framework specifies the population that benefits from water conversion, as well as the number of village health facilities, which are important indicators of public health and social welfare. The implementation of critical forestry programs illustrates the effectiveness of ecological restoration and resource management in rural areas, acting as a significant indicator of sustainable development capabilities.

The reaction is grouped into three categories: technological, ecological restoration, and societal. The technical response analyzes the usage level of biogas digesters, whereas the soil erosion control area reflects the efficiency of ecological restoration and land preservation, which is an important indicator of the ecological response. Natural disaster relief represents society’s ability to respond to disasters and is an important indicator of the effectiveness of the social response and disaster management efforts.

3.3. Research Methods

3.3.1. Entropy–TOPSIS Method

Entropy–TOPSIS quantitatively examines the total capacity of rural systems to absorb shocks (e.g., disasters and economic fluctuations) via objective assignments and distance ranking, thus providing a data-driven foundation for building resilience [41].

Prepare the original matrix $X$. The matrix $n\times m$, where n is the number of objects and m is the number of indicators, contains the rating data for different objects under different indicators.

$$X=\left[\begin{array}{cccc}{x}_{11}& x_{12}& \dots & x_{1m}\\ x_{21}& x_{22}& \dots & x_{2m}\\ ⋮& ⋮& \ddots & ⋮\\ x_{n1}& x_{n2}& \dots & x_{nm}\end{array}\right]$$

(1)

The original matrix is normalized and standardized to obtain a normalized matrix $r_{ij}$, where the range method is used.

Positive indicator:

$$r_{ij}={\displaystyle \frac{x_{ij}-x_j^{\min }}{x_j^{\max }-x_j^{\min }}}$$

(2)

Negative indicator:

$$r_{ij}={\displaystyle \frac{x_j^{\max }-x_{ij}}{x_j^{\max }-x_j^{\min }}}$$

(3)

where $r_{ij}$ is the $j$th indicator value for the $i$th sample.

Calculate the entropy value $e_j$ of the $j$ th indicator. When $P_{ij}=0$, $P_{ij}\ln P_{ij}=0.$

$$e_j=-{\displaystyle \frac{1}{\ln n}}\sum _{i=1}^n{P}_{ij}\mathrm{l}\mathrm{n}{P}_{ij},(j=\mathrm{1,2},\dots ,m)$$

(4)

Calculate the weight $W_j$ of the $j$ th indicator.

$$w_j={\displaystyle \frac{1-e_j}{\sum (1-e_j)}}$$

(5)

The smaller the entropy value ($e_j\to 0)$, the greater the variability of the indicator and the higher the weight.

The construction of weighted standardized matrices is as follows:

$$v_{ij}=w_j\times r_{ij}$$

(6)

Ideal solution:

$$V^{+}=\left\{max{v}_{ij}|j\in J^{+}\right\},V^{-}=\left\{min{v}_{ij}\right|j\in J^{+}\}$$

(7)

Calculate the Euclidean distance of each alternative to the positive ideal solution and the negative ideal solution:

$$D_i^{+}=\sqrt{\sum {\left(v_{ij}-V_j^{+}\right)}^2},D_i^{-}=\sqrt{\sum {\left(v_{ij}-V_j^{-}\right)}^2}$$

(8)

Calculate the relative closeness $C_i$ for each alternative:

$$C_i={\displaystyle \frac{D_i^{-}}{D_i^{+}+D_i^{-}}}$$

(9)

Here, $C_i$ranges from [0,1], and a higher value indicates a better alternative.

3.3.2. PLS-SEM

PLS-SEM (partial least squares structural equation modeling) is a multivariate statistical approach that combines partial least squares regression and structural equation modeling. PLS-SEM was preferred over CB-SEM for the following reasons: The dataset includes 15 indicators from four provinces over a six-year period, with a limited sample size. PLS-SEM just requires adherence to the “10-fold rule”, but CB-SEM often requires over 200 samples for consistent parameter estimates. The indices of rural resilience have a skewed distribution; PLS-SEM handles non-normal data via iterative weighting, whereas CB-SEM uses maximum likelihood estimation under the assumption of normality. PLS-SEM maximizes the endogenous variables, which improves structural equation modeling. Structural equation modeling (SEM) improves the predictive capacity by raising the explanatory variance of endogenous variables, making it suitable for exploratory research, whereas covariance-based SEM (CB-SEM) emphasizes theoretical validation and struggles to converge on complex models. PLS-SEM permits mean interpolation or the EM approach for dealing with missing values in the almanac data, whereas CB-SEM requires several interpolations and is sensitive to the existence of missing data. To summarize, PLS-SEM is more advantageous for measuring rural resilience when dealing with small sample sizes, non-normative data, and sophisticated models [42,43].

4. Results

4.1. Overall Assessment of the Level of the Rural Resilience Level

Table 2. Measurement and ranking of rural resilience in the Yangtze River Delta, 2012–2022.

City	Year	D	Rank	P	Rank	S	Rank	I	Rank	R	Rank
Shanghai	2012	0.05	6	0.36	5	0.47	5	0.41	3	0.16	6
	2014	0.17	5	0.19	6	0.45	6	0.25	6	0.23	5
	2016	0.81	2	0.43	4	0.83	1	0.63	1	0.87	1
	2018	0.90	1	0.59	3	0.71	2	0.31	5	0.35	4
	2020	0.34	4	0.90	1	0.53	4	0.34	4	0.37	3
	2022	0.42	3	0.85	2	0.54	3	0.59	2	0.40	2
Anhui	2012	0.41	6	0.24	5	0.34	6	0.10	6	0.96	1
	2014	0.43	5	0.24	6	0.81	1	0.11	5	0.37	5
	2016	0.48	4	0.26	4	0.37	4	0.47	3	0.50	4
	2018	0.55	3	0.49	3	0.50	3	0.41	4	0.34	6
	2020	0.57	2	0.61	2	0.34	5	0.67	2	0.57	2
	2022	0.59	1	0.93	1	0.53	2	0.75	1	0.51	3
Jiangsu	2012	0.54	1	0.00	6	0.31	6	0.64	2	1.00	1
	2014	0.52	2	0.50	4	0.33	5	0.69	1	0.16	6
	2016	0.47	3	0.27	5	0.48	3	0.50	4	0.16	5
	2018	0.46	4	0.73	2	0.55	2	0.53	3	0.24	4
	2020	0.37	6	0.53	3	0.88	1	0.42	5	0.29	2
	2022	0.46	5	0.87	1	0.46	4	0.27	6	0.27	3
Zhejiang	2012	0.62	1	0.06	6	0.65	2	0.42	4	0.64	2
	2014	0.57	2	0.38	5	0.85	1	0.42	3	0.66	1
	2016	0.40	3	0.53	3	0.54	4	0.29	6	0.36	6
	2018	0.35	6	0.44	4	0.57	3	0.37	5	0.36	5
	2020	0.36	5	0.79	2	0.53	5	0.70	1	0.43	4

shows the weights and ranks of the DPSIR dimensions in each province. Shanghai succeeded in both driving force and response, earning a good overall resilience rating. Between 2012 and 2022, Shanghai’s driving force ranking steadily rose from sixth to first place, owing mostly to its significant rural retail sales and rural population share, which highlight the rural regions’ strong economic vitality and demographic foundation. Shanghai’s ranking in the pressure dimension improved from fifth in 2012 to second in 2022, owing to significant advances in agricultural productivity and natural disaster prevention. The state dimension was undeniably remarkable, ranking first in 2016 and second in 2018, owing to large afforestation initiatives and the extensive coverage of natural reserves. Shanghai’s impact rating shifted from third in 2012 to second in 2022, indicating continued investment in rural public services and infrastructure development. The response component was the most noticeable, ranking first in 2016 and second in 2022, demonstrating its outstanding disaster management and environmental protection skills.

In Anhui, rural resilience levels have been steadily increasing. However, performance in the pressure component was more inconsistent. Between 2012 and 2022, An-hui’s driving force rating steadily rose from sixth to first place, indicating a strengthening of the rural economy. Anhui’s ranking on the pressure dimension ranged from fifth in 2012 to first in 2022, suggesting difficulties and hurdles in agricultural production and natural disaster management. The performance in the state dimension remained essentially steady, coming in first in 2014 thanks to considerable deforestation and the significant coverage of nature reserves. In the impact dimension, An-hui’s rating rose steadily from sixth in 2012 to first in 2022, suggesting significant progress in rural public services and infrastructure development. The response dimension showed greater volatility; after ranking first in 2012, it fell to third place in 2022, indicating a need for improvement in crisis management and environmental preservation.

Jiangsu had broadly consistent rural resilience levels, although with somewhat weak performance in the response component. Between 2012 and 2022, Jiangsu’s driving force position dropped from first to fifth, reflecting a decline in the rural economic basis. Jiangsu’s position in the pressure dimension improved from sixth in 2012 to first in 2022, suggesting significant progress in agricultural output and natural catastrophe management. Its performance in the state dimension has been continuously good, with second position in 2018 and first place in 2020 due to huge afforestation and an extensive coverage of natural reserves. Jiangsu’s impact rating fell from second in 2012 to sixth in 2022, showing that more efforts are needed to improve rural public services and infrastructure development. The response component declined from first position in 2012 to third place in 2022, indicating a need for enhanced disaster management and environmental protection.

Zhejiang displayed outstanding rural resilience, particularly in the impact component. Between 2012 and 2022, Zhejiang’s driving force position fell from first to fifth, suggesting a degradation in its rural economic basis. The pressure dimension’s ranking rose from sixth in 2012 to first in 2022, demonstrating significant advances in agriculture productivity and natural disaster management. The performance in the state dimension remained essentially stable, with a first-place ranking in 2014, as it was aided by vast afforestation and an extensive coverage of nature reserves. Zhejiang’s impact rating has steadily improved, rising from fourth in 2012 to first in 2022, signifying significant success in rural public services and infrastructure development. The response component showed substantial variability, ranking first in 2014 but falling to fourth place in 2022, indicating that further measures in disaster management and environmental conservation are necessary.

4.2. Overall Rural Resilience Level Trend

Figure 2 displays the general patterns of the DPSIR dimensions. The driving force (D) indicator showed continuously high values in most years, peaking at around 2.6 in 2022, suggesting the increased importance of socioeconomic activities in boosting regional resilience. Between 2012 and 2022, the driving force indicator showed a fluctuating rising trend, indicating a constant growth in the rural population and economic progress in the Yangtze River Delta area. This growing drive has provided a solid foundation for regional resilience. However, it may also pose risks such as resource exhaustion and environmental pressure.

The pressure (P) indicator has been very stable over the years, with values ranging from 1.8 to 2.0. This represents a broad balance between environmental and human stresses on regional resilience levels. Despite rapid economic growth in the Yangtze River Delta, environmental governance and policy management have prevented significant strain escalation.

From 2012 to 2022, the state (S) indicator showed small changes between 1.6 and 2.0, showing the health of regional ecosystems and resource stock levels.

The impact (I) index rose significantly in 2018 and 2020, reaching values of 2.2 and 2.4, respectively. These values show that environmental and social changes throughout these years had a significant impact on regional resilience. These oscillations can be attributed to factors such as significant climate disasters and economic structural changes.

The response (R) indicator peaked in 2016 and 2022, with values of 2.2 and 2.4, indicating that policies and governance measures implemented during these years significantly contributed to the improvement of regional resilience. The variations in the response indicator represent ongoing efforts in environmental governance, ecological preservation, and sustainable development in the Yangtze River Delta. Despite the efficiency of response measures in individual years, their aggregate values continuously fell short of those of the driving force indicator.

4.3. Analysis of Spatial Differences in Rural Resilience Level

Figure 3 depicts the regional and temporal trends in rural resilience levels across the YRD region from 2012 to 2022. It describes the complete process of natural resource development and environmental management in the YRD region. Between 2012 and 2022, the total rural resilience in the YRD region showed a fluctuating upward trend. However, each province’s performance varies significantly. In most years, Zhejiang and Jiangsu outperform Anhui and Shanghai in terms of rural resilience, demonstrating the effectiveness of their rural development programs and resource integration skills. Anhui and Shanghai’s rural resilience levels have fluctuated significantly in recent years, demonstrating that addressing natural restrictions and socioeconomic transitions remains tough. Regional disparities may be closely related to the level of economic development, natural resource availability, and the effectiveness of policy implementation in each province.

Figure 3a shows the geographical distribution for the year 2012. Zhejiang has the greatest level of rural resilience, followed by Jiangsu, while Anhui and Shanghai perform substantially better. This finding suggests that Zhejiang and Jiangsu began rural infrastructure development and ecological and environmental protection earlier, with significant consequences. Anhui and Shanghai may see reduced rural resilience as a result of economic structural instability or delayed policy execution.

Figure 3b shows the geographical distribution for the year 2014. Zhejiang remains the dominant area; however, Jiangsu’s rural resilience has significantly declined. Anhui and Shanghai both had a decrease in resilience levels, with Anhui hitting a low point, suggesting probable consequences from natural catastrophes or economic instability during the year. Anhui had catastrophic flooding in 2014, causing a significant reduction in rural resilience.

Figure 3c shows the geographical distribution for the year 2016. Anhui’s rural resilience level increased dramatically, making it the best-performing province that year, whereas Zhejiang’s and Jiangsu’s resilience levels decreased. This shift might be linked to increased investment in rural infrastructure development and environmental protection in Anhui. Concurrently, Zhejiang and Jiangsu may have seen a fall in resilience levels due to environmental stressors or political changes.

Figure 3d shows the geographical distribution for the year 2018. Anhui’s rural resilience has diminished, yet it remains better than that of Shanghai. Zhejiang and Jiangsu had decreased resilience, indicating that they may have faced more significant environmental restrictions or socioeconomic challenges this year. For example, Jiangsu may have felt the effects of its agricultural production model transformation in 2018, resulting in a lack of considerable increase in its rural resilience level.

Figure 3e shows the geographical distribution for the year 2020. Jiangsu’s rural resilience has significantly increased, approaching that of Zhejiang. Anhui’s and Shanghai’s resilience ratings are rather steady, indicating that the provinces are gradually fine-tuning their reactions to environmental pressures and socioeconomic changes. This change might be attributed to increased enforcement of disaster management and environmental conservation rules within the province.

Figure 3f depicts the geographical distribution for the year 2022. Zhejiang’s rural resilience level has increased once more, establishing it as the best-performing province for the year, while Jiangsu’s has decreased. The resilience levels in Anhui and Shanghai are generally constant, showing that the provinces are making progress in the development of rural resilience. Zhejiang’s dominance may be attributed to its continued investment in rural public services and infrastructure development.

4.4. PLS-SEM Evaluation Results

The major goal of PLS-SEM is to determine the authenticity and reliability of data. The dependability evaluation assesses the stability of the data and the model, including internal consistency and joint dependability. To be considered internally consistent, an entity’s Cronbach’s alpha must be more than 0.6 and its composite reliability (CR) must be at least 0.7. Validity assessments include both convergent and discriminant validity, which evaluate the measurement’s suitability. To establish convergent validity, the average variance extracted (AVE) must exceed 0.5, which means that the AVE accounts for more than half of the indicator’s variation. A higher AVE is associated with a greater chance of model correctness. To ensure discriminant validity, the square root of each average variance extracted (AVE) must be greater than the variable’s correlation with other variables across the diagonal. This study used Smart PLS 4.0 for data analysis, and the measurement model displayed significant composite reliability and discriminant validity, as shown in

Table 3. Reliability test results.

	D	I	P	R	S
Cronbach’s alpha	0.82	0.819	0.824	0.804	0.889
CR	0.895	0.893	0.898	0.885	0.931
AVE	0.746	0.737	0.75	0.722	0.819

and

Table 4. Measurement model discrimination effect detection.

	D	I	P	R	S
D	0.864
I	0.89	0.858
P	0.876	0.861	0.866
R	0.892	0.734	0.666	0.849
S	0.738	0.679	0.86	0.631	0.905

.

This study used the Smart PLS 4.0 software to determine the R2 value, hence assessing the model’s explanatory power. The R2 values of 0.796, 0.741, 0.767, and 0.739 indicate that the model has significant explanatory power. The driving variables, influencing variables, pressure variables, and state variables have R2 values of 0.791, 0.735, 0.761, and 0.733, indicating that they have significant explanatory power for the model.

The t-statistic value is calculated using the Bootstrapping process, which samples 5000 times at the 95% confidence range. At the 5% significance level, the hypothesis is considered statistically significant. When the t-value exceeds 1.96, the hypothesis is considered significant. The path coefficients and t-values show that hypotheses H1–H4 are accepted at the 5% significance level. This demonstrates that the causal relationships are represented in the model (Figure 4). The results show that the suggested index approach for assessing rural resilience is suitable for assessing the environmental circumstances of rural resilience in China.

This study re-evaluated the importance of the route coefficient to determine the relevance of the overall impact. Path relationships are classified into two types: direct path relationships, in which the path coefficient represents the direct effect between two variables, and indirect path relationships, in which one or more intermediary variables are used between the initial and terminal variables, with the indirect impact calculated by multiplying the respective direct path coefficients. The overall impact is calculated by adding the magnitudes of the direct impact and indirect effect. Table 3 depicts the effect channels for each conceivable variable. In contrast, Table 4 depicts particular indirect channels, allowing for a more in-depth investigation of the causative linkages within the DPSIR model for human settlements in the Yangtze River Delta’s rural districts.

Table 5. Path coefficient.

Hypothesis	Path	Coefficient	T	P	Outcomes
H1	D -> P	0.876	42.509	0	yes
H2	P -> I	0.861	23.413	0	yes
H3	P -> S	0.86	22.755	0	yes
H4	R -> D	0.892	24.459	0	yes

shows that the path coefficient for reaction to the driver is the highest, indicating that policy intervention has a significant direct impact on rural development. Biogas projects strengthen the energy framework, potentially lowering living costs and increasing rural consuming capacity (retail sales). Soil erosion control enhances arable land quality, indirectly raises agricultural revenue, and boosts rural economic viability. Natural disaster relief stabilizes communities’ livelihoods and reduces population migration, thereby sustaining the rural demographic share. In comparison to other elements, R has the greatest effect on D, indicating that policy reaction (R) is the key stimulant for modern rural development, rather than relying only on natural or economic underpinnings.

R has a significant positive direct influence on D and a positive indirect influence on P, and all of these effects are greater than the 1% significance threshold shown in

Table 6. Indirect path effects.

Path	Coefficient	T	P
D -> P -> S	0.753	21.145	0
R -> D -> P -> I	0.672	14	0
R -> D -> P	0.781	21.675	0
D -> P -> I	0.754	20.897	0

, implying that as the rural population and purchasing power grow, advances in modern science and technology occur, and social security improve. However, it also causes increased pollution in agriculture. D has a positive direct influence on P and an indirect influence on S, with a total impact of 0.753. This means that the rural population and economic expansion will increase ecological pressures in rural areas and have an impact on natural resources.

The influence of P on S is small, implying that the country’s proactive development of protected areas and afforestation has little impact on the growing levels of rural environmental pollution. As a result, further efforts are needed in terms of national policy. The influence of P on I is approximately equal to its impact on S, indicating that while the rural environment is gradually deteriorating, individuals’ living situations are improving; consequently, it is critical to reduce the harm to rural agriculture.

4.5. PLS-SEM Evaluation Results

Table 7. Final scores by province.

	D	I	P	R	S
Anhui	1.152	0.484	1.424	1.486	1.352
	1.09	0.527	1.373	0.088	1.928
	1.007	1.566	1.717	0.425	1.945
	0.934	1.28	0.928	0.201	1.238
	0.805	1.989	2.033	0.732	1.592
	0.821	2.301	0.815	0.573	−0.602
Shanghai	−1.712	−1.403	−1.25	−1.645	−0.994
	−1.709	−1.478	−1.264	−1.638	−0.996
	−1.625	−1.207	−1.278	−1.506	−0.98
	−1.618	−1.476	−1.29	−1.625	−0.983
	−1.694	−1.405	−1.302	−1.62	−0.986
	−1.65	−1.283	−1.31	−1.602	−0.987
Jiangsu	0.598	0.407	0.777	0.695	0.33
	0.534	0.397	0.606	0.044	0.228
	0.456	0.177	0.611	0.051	−0.251
	0.439	0.212	0.393	0.131	−0.06
	0.245	0.091	0.388	0.044	−0.208
	0.317	−0.08	0.237	−0.09	−0.552
Zhejiang	0.235	−0.219	−0.225	0.835	0.115
	0.261	−0.198	−0.329	0.999	−0.203
	0.273	−0.27	−0.47	0.555	−0.187
	0.283	−0.215	−0.5	0.574	−0.471
	0.169	0.169	−0.654	0.662	−0.484

shows significant variances in each province’s ratings across the DPSIR model dimensions, suggesting their varying success in economic development, environmental pressure, status, effect, and reaction. Anhui experiences strong economic development and policy implementation, but its environmental reaction is variable. Shanghai’s overall performance is weak, particularly in terms of aligning economic development with environmental concerns. Jiangsu has respectable results in economic development and environmental management; however, the effectiveness of its policy implementation is variable. Zhejiang excels in its environmental response but falls short in ecological pressure and economic driving force.

Anhui performs admirably on the D, P, and S dimensions, demonstrating strong skills in economic development, environmental pressure management, and reaction tactics. Nonetheless, the dimension is highly variable, with specific values being quite low, showing differences in ecological governance and response tactics. The greatest score in the R dimension is 1.486, while the minimum is 0.088, showing a considerable difference. Furthermore, Anhui’s continuously high scores in the S dimension indicate that it has been rather effective in policy reaction and implementation.

Shanghai has low ratings in all five areas, indicating major issues in environmental governance and sustainable development. Specifically, the lowest scores in the D and R dimensions are −1.712 and −1.645, indicating a lack of congruence between economic growth and environmental reactions. Despite Shanghai’s low S dimension ratings, they are somewhat higher (maximum of −0.98), indicating some attempts at policy remedies, but with limited overall efficacy.

Jiangsu has largely positive outcomes in the D, P, and R domains, indicating strong success in economic development and environmental responsiveness. Nonetheless, the I and S dimensions had lower scores, including some negative values, indicating gaps in environmental impact management and policy responses. The minimal score in the S dimension is −0.552, indicating the unpredictability of policy implementation effectiveness. Jiangsu’s D dimension scores are rather stable, with a high score of 0.598, suggesting a strong economic momentum.

Zhejiang excels in the R dimension, which indicates robust environmental reaction capabilities. Nonetheless, its ratings in the I and P categories are primarily negative, highlighting challenges related to the environmental effect and pressure. The minimum score on the P dimension is −0.654, suggesting significant environmental difficulties. Although Zhejiang’s D dimension ratings are relatively stable, the overall level remains low, with a maximum score of only 0.283, suggesting a weak economic driving force.

5. Discussion

5.1. Overall Assessment of Trends in Rural Resilience Levels

Rural revival is a significant development initiative in China; however, the evaluation of rural ecological revitalization via the indicator system is still in its early stages, and each province’s developmental status is uncertain. This study uses the DPSIR framework to develop an assessment system consisting of five basic indicators, eleven secondary indicators, and fifteen tertiary indicators. Weights were then applied to the indicators, and the provinces were arranged in the order of the five DPSIR criteria. To summarize, the changes in rural resilience levels from 2012 and 2022 are as follows: the YRD region had a decline from 2012 to 2016, a climb beginning in 2018, and eventually achieved a condition of stability.

Factors leading to a drop in resilience levels between 2012 and 2016: Initially, economic restructuring exerts specific forces. Villages in the Yangtze River Delta are facing difficulties such as worker outmigration and agricultural decline as a result of growing urbanization. Resource-dependent regions, like northern Jiangsu and northern Anhui, are less resilient to disasters due to their monolithic industrial infrastructure. Second, there is a vacuum in policy convergence. Prior to the implementation of the rural revitalization plan in 2017, rural development lacked comprehensive governmental support. According to the research, low investment in rural infrastructure and public services before 2016 reduced the ability to withstand natural disasters (e.g., floods and typhoons) and socioeconomic shocks, ultimately leading to ecological decline and environmental pollution. Pollution from industrialization in the Yangtze River Delta has reduced ecological resilience, with certain places in northern Jiangsu Province experiencing a drop in the resilience index in 2016 due to ecological inadequacies.

Principal variables leading to resilience strengthening after 2018: The Strategic Plan for Rural Revitalization (2018–2022) strengthened the YRD’s economic and social resilience ecosystem through sector integration and infrastructure improvements. Since 2018, ecological governance has demonstrated effectiveness. Jiangsu and other regions have significantly improved their environmental resilience through pollution control (e.g., increased sewage treatment rates) and ecological restoration (e.g., soil and water erosion management), resulting in a resurgence of the Jiangsu comprehensive resilience index by 2020. Multidimensional governance gradually improved procedures. The merger of social organizations and the strengthening of farmers’ agencies have strengthened the rural organization’s ability to manage unanticipated threats [44].

Underlying variables that have contributed to stability in recent years after 2018 include the regional synergistic effect. The Yangtze River Delta’s integration policies, such as the 2019 Outline of the Plan for the Integrated Development of the Yangtze River Delta Region, have improved the exchange of urban and rural resources, reduced the income disparity between Southern Jiangsu and Northern Anhui, and achieved a state of dynamic equilibrium overall [45], thereby structuring resilience capacity. Increases in economic resilience (e.g., a larger share of non-agricultural sectors) and social resilience (e.g., investments in healthcare and education) have resulted in a higher threshold of resistance to disruptions and fewer swings.

This trend represents the transition of the YRD’s countryside from a “passive shock response” to “active adaptive restructuring”, with policy interventions (e.g., rural revitalization), technological innovation (e.g., green patents), and spatial restructuring (e.g., county economic synergy) as the primary variables. To strengthen resilience, it is critical to address shortcomings in innovation and transformation capabilities, such as the adoption of agricultural science and technology.

5.2. The Intrinsic Link Between Levels of Rural Resilience

The linkages inside the DPSIR framework are investigated by merging it with PLS-SEM based on its indicators. The path coefficients show that a mix of “drivers” from rural polices, “responses”, and environmental “pressures” causes variations in the amount of rural resilience.

This outcome might be attributed to a variety of circumstances. Policy responses have a dual transmission influence. The YRD’s policy reaction boosts economic momentum through financial investments and industry aid, directly increasing economic resilience. The dynamic interaction between driving force and pressure is shown. While economic incentive promotes industrial growth, it also increases resource consumption and pollutant emissions. The PLS-SEM path analysis shows that policy responses reshape the connection between these variables through technological intervention and institutional innovation. A comprehensive reconfiguration of urban–rural integration is needed. The policy strategy promotes the mobility of resources between urban and rural areas through factor marketization reforms, thereby dismantling the traditional dual structure. Representative examples show how the three southwestern districts of Shanghai reorganized the “three living spaces” through considerable land improvement, turning environmental difficulties into a catalyst for spatial optimization. This systematic shift has resulted in a U-shaped curve of “initial decline, followed by recovery” in the resilience of the Yangtze River Delta’s rural communities [46].

The final PLS-SEM scores show a discrepancy in the assessment values and level of change across each DPSIR subsystem, with the “impact” dimension having the highest evaluation value, followed by “pressure” and “state”. In the DPSIR causal chain, the “im-pact” dimension (I) serves as the last link, emphasizing the cumulative influence of the preceding subsystems. PLS-SEM route coefficients show that the Yangtze River Delta area, via industrial upgrading and eco-compensation policies (R), directly improves the effect indicators of livelihoods, health, and income levels. By observing the delay in pressure–state dynamics, we find that “Pressure” (P) and “state” (S), as intermediary transmission variables, have temporal and spatial delays in data collecting. The state indicator requires extended observation to identify changes, and the study shows that its score has a marginally growing trend alone after 2019. The monitoring delay causes a considerable measurement error of the two in the PLS-SEM model, lowering the assessment value [47].

5.3. Policy Recommendations

China has made significant advances in ecological rejuvenation. The environment has been significantly improved as part of the rural revitalization process, resulting in a gradual improvement in the quality of life in rural communities. Nonetheless, disparities in rural development have resulted in major differences in developmental levels among provinces and regions. To strengthen China’s overall resilience and achieve SDGs 7 (clean water and sanitation), 8 (decent work and economic growth), and 11 (sustainable cities and communities), it is critical to identify priority policy targets, regional development weaknesses, and spatial aggregation and spillover effects. This will help with the optimization of strategic resource allocation. This research makes the following policy recommendations.

• Given the relative weights of the DPSIR in the four provinces, it is advised that tailored solutions be applied in specific regions, particularly in areas with fewer rural populations and limited land availability, like Shanghai and other heavily urbanized areas. The government must maintain strict control and implement focused programs to enhance the rural environment. In underdeveloped areas such as Anhui, there is a significant mismatch between environmental issues and government responses. A dynamic matching system for “pressure-response” should be developed, with a focus on the progress of digitally enabled ecological compensation schemes. This will increase the social awareness of environmental protection and strengthen the ability for responding.
• Based on the path coefficient results from the PLS-SEM analysis, it is recommended to implement a resilience enhancement mechanism consisting of “intelligent monitoring—dynamic compensation—digital governance”, the real-time monitoring of environmental pressure indicators via the Internet of Things (IoT), the automatic activation of eco-compensation when pressure values exceed standards, and a precise allocation of governance resources using digital technology. A digital resilience platform should also be built to promote collaboration across the three provinces and one city, with the goal of simulating policy impacts using digital twin technologies. The method focuses on optimizing a mix of environment-based (40%), demand-based (30%), and supply-based (30%) policy instruments in order to achieve a dynamic equilibrium between the transformation of driving forces and pressure alleviation, thereby systematically increasing rural resilience. The strategy can shorten the policy lag by 2.1 years while increasing response efficiency by more than 40%.
• Based on the final evaluation ratings, it is recommended to develop steps to strengthen resilience along the causal chain. Anhui has the least substantial effect (S) and may link the drivers to improve ecological restoration, disaster prevention, and management [48]. Shanghai has the most insufficient drivers and requires the establishment of population policies and transformational pressure to improve its environmental quality [49]. Jiangsu ® is the least resilient; it is needed to strengthen technology solutions, consolidate funds to support these measures, and develop cross-domain collaboration to collectively manage the Yangtze River Basin’s environmental issues [50]. Zhejiang province (S) is the least favorable and needs to improve its pressure regulation and implement intelligent governance [51].

6. Conclusions

6.1. Main Contributions

This study uses the DPSIR framework and the PLS-SEM system to thematically examine rural resilience in China’s Yangtze River Delta region, elucidating internal logical links and spatiotemporal developmental patterns. This study combines the entropy weight technique with TOPSIS to examine the environmental assessment of rural resilience levels, creates an evaluation index system, and quantitatively analyses the causal links within the DPSIR models. Second, this study uses DPSIR and PLS-SEM to show that the rural socioeconomic “response level” to changes in rural “driving forces” and environmental “pressures” determines the level of rural resilience. Together, they influence the degree of rural resilience. The linkages between the model’s five aspects are extensively examined; essential factors and limits are identified, and specific recommendations are made. The key findings reveal that (1) rural resilience in the Yangtze River Delta region decreased from 2012 to 2016, then increased in 2018, and finally stabilized. The “driving force” arising from the rural policy “response” and environmental “pressure” has jointly produced changes in the level of rural resilience. (3) The assessment values and the level of change differ across each subsystem, with the “impact” dimension having the highest evaluation value, followed by “pressure” and “status”. This study evaluates rural resilience and its driving mechanisms in the Yangtze River Delta region, providing a scientific foundation for optimizing resource allocation and policy interventions, thereby directly facilitating the achievement of SDG 7 (Improvement of Rural Environment and Sanitation), SDG 8 (Promotion of Rural Economic Resilience through Digital Governance), and SDG 11 (Enhancement of Urban and Rural Sustainable Development Capability through Regional Linkages).

6.2. Research Significance

This study uses the DPSIR theoretical framework and the nonlinear PLS-SEM model to investigate the dynamic coupling processes of five subsystems within the Yangtze River Delta’s rural resilience system. The study made three significant methodological advances: using the partial least squares algorithm to identify nonlinear interactions, introducing a time-varying pathway system to quantify dynamic evolution, and developing a coupling algorithm of entropy weight–TOPSIS and PLS-SEM for weight adaptation. These advances not only established a more precise rural resilience evaluation system, but they also identified significant shift features in the rural resilience driving mechanism around 2016. These findings provide unique theoretical insights and analytical tools for understanding the complexities of rural socio-ecological systems.

The study’s findings provide cohesive solutions for simultaneously advancing several sustainable development goals: improving rural water supply and sanitation through intelligent monitoring systems (SDG7), optimizing rural industrial structures through digital governance platforms (SDG8), and promoting community sustainability through regional synergistic mechanisms (SDG11). The proposed dual-wheel drive model of “technological innovation and institutional synergy”, which achieves coordinated governance across three provinces and one city through the digital resilience platform, addresses the issue of uneven rural development and provides a replicable Chinese experience for localizing global sustainable development goals. This multi-objective synergy research paradigm emphasizes the crucial importance of increasing rural resilience in order to achieve the 2030 Agenda.

6.3. Limitations and Future Research

This study has certain limitations. The lack of indicators that adequately depict the current state of rural regions may lead to mistakes caused by data availability limits. This study’s duration is insufficient to identify long-term trends and changes, and there is opportunity to improve the actual values of the PLS-SEM model. It is recommended to integrate additional variables that reflect rural-specific characteristics and extend the study’s temporal scope to more comprehensively assess changes in rural village resilience, thereby permitting more relevant results in future research.